REVIEW Open Access

Assessing value of innovative moleculardiagnostic tests in the concept of predictive,preventive, and personalized medicineIldar Akhmetov1 and Rostyslav V. Bubnov2,3*

Abstract

Molecular diagnostic tests drive the scientific and technological uplift in the field of predictive, preventive, andpersonalized medicine offering invaluable clinical and socioeconomic benefits to the key stakeholders. Although theresults of diagnostic tests are immensely influential, molecular diagnostic tests (MDx) are still grudgingly reimbursed bypayers and amount for less than 5 % of the overall healthcare costs. This paper aims at defining the value of moleculardiagnostic test and outlining the most important components of “value” from miscellaneous assessment frameworks,which go beyond accuracy and feasibility and impact the clinical adoption, informing healthcare resource allocationdecisions. The authors suggest that the industry should facilitate discussions with various stakeholders throughout theentire assessment process in order to arrive at a consensus about the depth of evidence required for positivemarketing authorization or reimbursement decisions. In light of the evolving “value-based healthcare” deliverypractices, it is also recommended to account for social and ethical parameters of value, since these are anticipated tobecome as critical for reimbursement decisions and test acceptance as economic and clinical criteria.

Keywords: Predictive Preventive Personalized Medicine, Market access, Value, Strategy, Companion diagnostics, Cost-effectiveness, Reimbursement, Health technology assessment, Economic models

ReviewThe value-based approach is needed in economic modelsin predictive, preventive, and personalized medicineFundamental shifts in the healthcare paradigm, drivenby the gradual transition to “patient-centered” value-basedhealth service delivery, open new horizons to personalizedmedicine offering the potential to provide timely andcost-effective medical solutions to stratified patientsubpopulations with predictable outcome margins. Byusing biomarkers as measurable indicators of predis-position to or severity of a disease state, personalizedmedicine helps in early detection, monitoring, assess-ment of risks associated with a disease, and guidingtherapeutic decisions [1, 2].Today, close to 50 % of the early-stage pipeline assets and

30 % of late-stage molecular entities of the pharmaceutical

companies involve the use of specific biomarkers [3]. More-over, molecular diagnostic tests (also molecular genetic test-ing or MDx), which drive the growth in personalizedmedicine by detecting and measuring proteins, nucleicacids, or metabolites variations, represent the fastest devel-oping segment in the diagnostic (Dx) market enjoyinghealthy 10 % annual growth and promising to achieve USD12.78 billion by 2018 [4].With more than 26,000 diagnostic tests available for

over 3600 genes in the USA alone, personalized medi-cine undoubtedly becomes one of the hottest areas inthe global healthcare sector [5]. Yet, acceptance andadoption of personalized medicine by wider communitywithin the global health system requires that all majorstakeholders understand and are able to measure valueoffered by this practice, which is a herculean task con-sidering the diversity of stakeholders’ needs, lack of auniform assessment criteria, and often intangible multi-parameter nature of “value.”Unlike therapeutics (Rx), which undergo three phases

of clinical trials prior to marketing authorization and

* Correspondence: rostbubnov@gmail.com2Clinical Hospital “Pheophania” of State Affairs Department, Zabolotny Str.,21, Kyiv 03680, Ukraine3Zabolotny Institute of Microbiology and Virology, National Academy ofSciences of Ukraine, Zabolotny Str., 154, Kyiv 03680, UkraineFull list of author information is available at the end of the article

© 2015 Akhmetov and Bubnov. Open Access This article is distributed under the terms of the Creative Commons Attribution4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 DOI 10.1186/s13167-015-0041-3

whose effect on patients can be quite straightforwardlydemonstrated by patient-reported outcome measures,there is no clarity in the molecular diagnostics field onhow much evidence is required to prove the value of atest. As a rule, most of diagnostic studies focus on accur-acy and feasibility, which is a hardly enough prerequisitefor better patient health or other downstream improve-ments [6, 7].Having no direct evidence on patient outcomes, a

diagnostic test is frequently looked upon as a non-valueadding service, since it does not treat patients like atherapeutic but rather improves physicians’ decisions ona therapeutic intervention. Nevertheless, it is estimatedthat the results of diagnostic tests are immensely influ-ential affecting around 60–70 % of all clinical decisions,although they still amount for only 4–5 % of healthcarecosts [8, 9]. Such underestimation of diagnostic valueleads to confined market access to safer and more effectivetherapies, poorer (often “non-value-based”) reimburse-ment, weaker intellectual property rights protection,underdeveloped incentivization systems for diagnostic in-novations, as well as limited returns on investment vis-à-vis high R&D costs.The purpose for this paper is to define the value of a

molecular diagnostic test and to outline the mostimportant components of value from miscellaneous as-sessment frameworks, which go beyond accuracy andfeasibility, informing healthcare resource allocation deci-sions and facilitating the clinical adoption.

Defining value of MDxMolecular diagnostic testing has moved to the forefrontof present-day clinical practice and is widely used inmiscellaneous areas of healthcare including oncology,neuroscience, cardiology, infectious diseases, metabolicdisorders, etc. The emergence of highly accurate anduser-friendly diagnostic tools based on blood, cerebro-spinal fluid, urine, and saliva, in addition to biopsied tis-sue, contributes to better acceptance and adoption ofMDx in physician offices, clinics, and community healthcenters that recognize value in individualizing diagnosisand treatment [10, 11].Fundamental shift in healthcare delivery is a harbinger

of new evolving diagnostic testing needs, which forcethe already sophisticated and sensitive assays to developinto more portable, easy-to-use, cost-effective and lesstime-consuming platforms. Thus, the recently approvedOraQuick ADVANCE™ test for a simple at-home use al-lows HIV patients to learn their status in just 20–40 minwith 99.87 % specificity and 93.64 % sensitivity, clearlydemonstrating the gradual transition from “laboratory-centered” to “user-centered” healthcare delivery [12].The other study of myoglobin and troponin point-of-care testing showed 55–66 % decrease in time to obtain

test results with 96.9 % sensitivity and 99.6 % NPV, whencompared to central laboratory procedures, suggestingthat the use of in vitro diagnostic kits in point-of-caretesting in certain cases may be more time-saving and ef-fective than conventional diagnostic approaches [13, 14].It is also worth noting that rapid diagnostic testing is

evidenced to reduce complications and avoid deathsamong patients. For example, the data from three ran-domized controlled trials (RCT) carried out in Denmarkdemonstrated reduction in the relative risk of deathfrom colorectal cancer (CRC) to less than 0.70 for sub-jects adhering to the screening [15]. Additionally, theuse of AdvanDx’s PNA FISH™ testing demonstrated a42 % reduction in mortality of patients with highly drug-resistant enterococcus faecium infections [16].The studies of clinical and economic impact of MDx

in the context of personalized medicine demonstrateunivocal results that signify value of individualized ap-proach to diagnosis and treatment. The report by AEI-Brooking Joint Center for Regulatory Studies projectedthat testing for variants in the CYP2C9 and VKORC1genes guiding the initial dosing of warfarin could pro-vide USD 1.1 billion in annual savings to the US health-care system and prevent 17,000 strokes with 85,000bleeding events [17]. Additionally, USD 604 million inannual savings would be realized if genetic tests limitinganti-epidermal growth factor receptor (EGFR) therapy tometastatic colorectal cancer patients with wild-typeKRAS tumors are carried out [18].Molecular diagnostic tests aid in guiding physicians’

decisions on treatment, which becomes even more top-ical in light of increasing heterogeneity of diseases, highdrug attrition, and associated problems of over- andundertreatment, tightening up budgets of healthcarecenters. For instance, genetic testing of breast cancerpatients with Oncotype DX™ assay resulted in alteredtreatment of 44 % of individuals, demonstrating high ac-curacy and bringing high degree of certainty of the testvalue [19]. Along the same lines, there is evidence thatdiagnostic tests have contributed to 30–50 % reductionsin direct hospital and outpatient charges by identifyingkey alterations in health status and facilitating modifica-tions in therapeutic interventions to improve patientoutcomes [8].MDx also improve adherence, compliance, and will-

ingness to undergo treatment or prevention by means ofbetter prognosis of disease occurrence and prediction ofthe response to treatment.Thus, PreDx Diabetes Risk™testing assesses the patient risk for getting type-2 dia-betes over a 5-year period motivating patients to under-take preventive measures and to change their unhealthylifestyles [20]. Similarly, the literature review of patientcompliance based on genetic testing in breast and colorec-tal cancers, hamochromatosis, thrombophilia, smoking

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 2 of 12

cessation, and obesity showed high increase in compliancerates associated with MDx [21].High value of molecular diagnostic testing is observed

in the pharmaceutical pipeline, as it facilitates discoveryof biomarker-based therapies targeting disease causes in-stead of symptoms. It is evident that today, 50 % of allclinical trials conducted by pharmaceutical companiescollect DNA from patients in order to facilitate bio-marker development [22]. Moreover, is also known thatbiomarker-based diagnostics used in clinical trials canincrease chances of regulatory approval [23, 24] and en-hance prescription [25, 26].In certain cases, MDx assist pharmaceutical companies

to rescue licensing and reimbursement by providingmore convincing arguments to regulatory authoritiesand insurers on clinical outcomes and cost-effectivenessthrough stratification of population. For example, Iressa®(gefitinib) withdrawn from the market after failing toprove survival benefit in phase 3 trials regained market-ing authorization in Europe in combination with theEGFR mutation test [27]. Likewise, Herceptin® (trastuzu-mab) was not considered to be cost-effective by NICEand SMC initially in the large gastric cancer population;however, once the HER2 overexpression subgroup wasdefined, the decision has been changed to positive [28].Finally, with advancements in bioinformatics and

emergence of novel multi-omic databases, like Open-BEL, it is anticipated that systems diagnostics (SysDx),incorporating a wide series of biomarkers from differentbiological disciplines, can add substantial clinical and so-cioeconomic value by being easily scalable to addressmuch larger groups of patients and by comprehensivelybreaking down a single complex disease into multipletargets with tailored treatment options.To sum up, molecular diagnostics go far beyond trivial

diagnosis, but instead are critical in identifying individ-ual risk for developing a disease, prescribing safe and ef-fective therapies, assessing response to a therapeuticintervention throughout the course of treatment, prepar-ing viable disease management strategies, etc. Moreover,new generation MDx can add downstream value byvirtue of their evolving characteristics such as greater ac-curacy, higher throughput, shorter testing time, simpli-city, portability, cost-effectiveness, and so on.

Assessing value of MDxIn view of the polysemic nature of “value” of MDx, thereare numerous interpretations and ways to measure thisconcept depending on which model is used and fromwhose perspective the technology is being assessed. Forinstance, the “innovativeness” of MDx is likely to be ofhigh value from a commercial standpoint but of lesservalue to patients, physicians, or payers. Along the samelines, the benefit of an accurate diagnosis may be of high

value to patients and doctors, although challenging toquantify and to validate for payers.

General value assessment frameworksPorter’s Value-Based Healthcare (VBH) model placespatient atop of the hierarchical pyramid of significance,arguing that the true value of any healthcare service,including diagnostic testing, can only be conceivedthrough a prism of the total bundle of products andservices delivered to an individual patient over a cycle ofcare by correlating the final patient outcomes with theassociated costs [29]. In other words, the only way toadequately assess the real value of a diagnostic test is toconsider it as an integral part of the combined “efforts”of all links in the value chain involved in the process ofdelivering value to a patient (e.g., healthcare providers,caregivers, manufacturers, pharmacists, and laboratories),assigning weights to each link based on its contribution tothe overall patient outcome and measuring the total costsrequired to deliver this outcome:

Value ¼ Patient outcomes�

Costs of delivering the outcomes

*The patient outcomes in this model include preven-tion of illness; early detection; right diagnosis; righttreatment to the right patient; rapid cycle time of diag-nosis and treatment; treatment earlier in the causalchain of disease; less invasive treatment methods; fewercomplications; fewer mistakes and repeats in treatment;faster recovery; more complete recovery; greater func-tionality and less need for long-term care; fewer recur-rences, relapses, flare ups, or acute episodes; reducedneed for ER visits; slower disease progression; less care-induced illness; etc. [30].Despite implementation difficulties, which are caused

by the need in substantial structural changes in health-care systems and stakeholders’ willingness to collaborateand to step out from the comfort zone, this model hasalready been accepted by a number of American healthcenters (e.g., Cleveland Clinic [31], Joslin DiabetesCenter [30], and Catawba Valley Medical Center [32])and won recognition of a large insurance association inthe USA (e.g., the Blue Cross Blue Shield of Michigan’sValue-Based Contract model [33]). The emergence ofnovel VBH practices, such as accountable care organiza-tions (ACOs) and patient-centered medical centers(PCMCs), may lead to new levels of data collection cap-turing value of diagnostic tests in conjunction with mis-cellaneous treatment options, admissions, surgeries, andother healthcare services [34].Another approach to assess value of a diagnostic test

was introduced by Fryback and Thornbury [35], whoproposed a hierarchical six-level evidence framework of

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 3 of 12

evaluating diagnostic technologies, later developed bySilverstein and Boland [36], Mackenzie and Dixon [37],Pearson et al. [38], and di Ruffano et al. [6], includingthe following:

▪ Technical characteristics (e.g., testing time,portability, ease to interpret results)▪ Diagnostic accuracy (e.g., sensitivity, specificity, areaunder the receiver operating characteristic curve)▪ Impact on diagnostic thinking (e.g., % of cases inwhich a physician is confident that the test changes thediagnosis)▪ Impact on therapeutic actions (e.g., % of cases inwhich the choice of treatment is changed afterinformation from the test is provided, shorter time totreatment, improved adherence/willingness to take Rx)▪ Impact on patient outcomes (e.g., differences inmortality, morbidity, or quality of life between patientsmanaged with the test and those managed without it,decreased re-hospitalizations)▪ Impact on societal outcomes (e.g., cost-effectivenessof an improvement in patient outcomes, such as costper life-year saved, calculated from a societal perspec-tive) [6, 35-38]

Harris et al. [39] complemented the above-mentionedframework outlining the necessity in measuring adverseeffects or changes in their incidence related to diagnos-tics and the associated therapeutics [39]. Additionally,the scholars argue that both immediate patient out-comes (e.g., blood pressure, glucose levels, or malignantneoplasm regression) and overall health outcomes (e.g.,incidence of heart attacks, diabetes progression, or can-cer survival) should be reflected in the assessmentmodel because surrogate markers alone (e.g., reductionin protein levels) may not be enough criteria to justifyreimbursement or regulatory approval.Apart from the criteria specified by Fryback and

Thornbury [35] and their successors, Ferrante di Ruf-fano et al. [6] also recommended using two additionalparameters to assess value—feasibility (e.g., acceptabil-ity, clinical contradictions, and failure rates) and testprocess (e.g., procedural harms or benefits, placeboeffect) [6].According to Harris et al. [39], Trikalinos et al. [40],

and Ferrante di Ruffano et al. [6], one of the most im-portant ways to provide real evidence for value of an in-novative diagnostic test is to either conduct RCTs or toconstruct decision analysis models that combine suffi-cient sources of evidence [6, 39, 40]. Considering thehigh cost of direct studies, the scholars advise to validateintermediate outcomes or surrogate markers as predictiveof patient outcomes, which are sufficient in many casesfor third-party payers to provide reimbursement.

The value assessment framework developed by the“Grading Quality of Evidence and Strength of Recommen-dations for Diagnostic Tests and Strategies” (GRADE)Working Group [41] also encourages the use of observa-tional studies to compare alternative diagnostic strategiesand assess direct patient-important outcomes in thosecases, where RCTs are not feasible to implement [41]. TheGRADE Working Group advices that the data set shouldprovide enough evidence on decreasing false negatives orfalse positives, increasing true positives or true negatives,high accuracy of patient classification vis-à-vis alternatetests, and improved outcomes of both affected and healthypatients. Moreover, if there is no available effective treat-ment linked to the test, the accurate diagnostic assay maystill be beneficial, once it reduces test-related side effectsand anxiety by avoiding wrong diagnosis or may contrib-ute to prevention and healthy lifestyle of patients bymeans of prognostic or predictive information.Hayes et al. [42] and Simon et al. [43] provide clear

guidelines on how to select efficient indirect “prospect-ive–retrospective” designs based on archived specimensfor establishing the medical utility of a prognostic orpredictive biomarker as an alternative to costly RCTs.These scholars proposed five levels of evidence (LOE) todetermine the clinical utility of a tumor marker and toreport the results of marker research, depending on thestudy design (A—“prospective,” B—“prospective with ar-chived specimen,” C—“prospective/observational” andD—“retrospective/observational”) and the variables in-volved (e.g., trial type, patient data, collection and stor-age of specimen, and statistical analysis and validation)[42-44].Building up on Wald and Cuckle [45] framework and

using terminology introduced by the Secretary’sAdvisory Committee on Genetic Testing, Haddow andPalomaki [46] developed the ACCE model to evaluategenomic tests based on 44 targeted questions groupedinto five categories: (1) clinical disorder and setting inwhich DNA testing is to be applied, (2) analytic validity,(3) clinical validity, (4) clinical utility, and (5) ethical,legal, and social implications (Fig. 1) [46]. This modeldiffers from the aforementioned frameworks by addressingseveral important contextual issues impacting value of aMDx, like availability of facilities and personnel, robust-ness of education materials, methods for long-termmonitoring, discrimination/ stigmatization, privacy, legalconcerns, consent, ownership of data and/or samples, pat-ents, licensing, proprietary testing, obligation to disclose,or reporting requirements, etc. [46].Haddow and Palomaki [46] and National Academy of

Sciences [44] further clarify that clinical validity is theassociation of a test result with an outcome (e.g., identi-fying disease and predicting adverse events) expressedby sensitivity, specificity, predictive values, odds and risk

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 4 of 12

ratios, and logistic regression analyses [44]. Analyticvalidity deals with the technical performance of the tes-t—its accuracy, repeatability, and reliability over time orunder the influence of external substances. As to clinicalutility, it describes how the test affects patient manage-ment and outcomes vis-à-vis the usual care and thuselucidates the significance of the test for individualpatient decision-making. While clinical and analyticvalidity is most of the time required for marketingauthorization, it is a clinical utility that plays a key rolein the vast majority of reimbursement decisions.Addressing the health outcomes issue, Bossuyt and

McCaffery [47] have introduced a broader classificationof patient outcomes that unlike Fryback’s and Thorn-bury’s [35] parameters incorporates six groups of testing“effects”—clinical management effect, direct health ef-fect, emotional effect, social effect, cognitive effect, andbehavioral response to testing [47]. Lee et al. [48] andKopits et al. [49] supported such a broad classification cit-ing examples from the Alzheimer’s studies, where patientsexpressed a willingness to know early in life about thechances of contracting the disease recognizing highemotional and social value in long-term planning (e.g.,choices on lifestyle, work, retirement, financial plans, andreproduction) [48, 49].According to Bossuyt and McCaffery [47], such multi-

variate taxonomy of patient outcomes broadens data

collection, helps to better assess value, and facilitates im-provements in overall delivery of quality healthcare.However, considering time and resource limitation, itcan hardly happen that a diagnostic company incorpo-rates all of these parameters into a value proof case. Forthat reason, a timely engagement with stakeholdersthroughout the entire assessment process is a prerequis-ite to better comprehension of stakeholders’ expecta-tions and maximization of the evidence quality.Finally, according to Task Force on Community Pre-

ventive Services, one of the most important consider-ations in assessing value of a diagnostic test is theperspective from which the analysis is carried out thatpredefines the composition of costs and health outcomesincluded in the analysis [50]. For example, a study con-ducted from a government perspective is likely to ac-count for only those costs and benefits experienced bythe government and disregard costs or benefits relevantto a health insurance purchaser, while value assessmentfrom patient perspective is likely to incorporate muchbroader variety of costs (Table 1).The table suggests that while the societal and patient

perspectives incorporate most of costs associated with adiagnostic test, payers’ and employers’ perspectives aremuch more selective. This proves the aforementionedstatement that timely engagement with stakeholders mayallow diagnostic companies to prepare more informed

Fig. 1 Graphic representation of the ACCE system for assessing genomic tests (adapted from Haddow and Palomaki [46])

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 5 of 12

and custom-made evidence on financial impact of Dxtesting with lesser time and money spent.To sum up, generating direct and indirect evidence of

causal relationships between a conducted test, improvedhealth outcomes, and the associated costs is a convin-cing way to prove value of diagnostics. However, consid-ering often nonlinear value chain of healthcare delivery,conducting such studies may be complicated, time-consuming, and costly. In view of the multi-parameternature of value, it is recommended thereby to facilitatediscussions with various stakeholders throughout the as-sessment process in order to arrive at a consensus aboutthe depth of evidence required for positive regulatory orreimbursement decisions.

Private payer perspectiveThe growing influence of payers and the need of diagnos-tic producers to dispel payers’ “value-for-money” concernhave endued the process of quantifying costs and out-comes in assessing value of molecular diagnostic testswith even greater importance. Currently, there is no uni-form method to measure economic value among privatepayers or state programs neither in the developed nor infast-growing economies, which leads to a continuous dis-connect in valuation of MDx in the eyes of payers anddiagnostic companies.Of traditional methods to assess economic value of

diagnostic tests comparing costs and outcomes, thereare two most widely used approaches—cost-effectivenessanalysis or CEA (expressed in delta costs compared withdelta health outcomes) and cost-benefit analysis or CBA(usually expressed purely in monetary terms). Addition-ally, cost-utility analysis (CUA), which is a subtype ofCEA, measures all costs in monetary units and benefitsin terms of quality-adjusted life years (QALYs). WhileCEA and CUA provide information as to whether ahealth technology maximizes the health of a populationin the resource-limited setting (widely used by payers),CBA is rather focusing on maximization of social welfarein light of societal budget constraints [51]. Yet, theaccuracy of either economic assessment method dependsvery much on what data is available, what assumptions

researchers make, how cost/outcome parameters areselected and validated, etc.In the recent survey among small and medium busi-

nesses dealing with health technology assessment(HTA), the interviewed firms accepted that the mainimpediments in the reimbursement process for theirproducts were poor understanding of specific payer re-quirements, insufficient scientific advices from the HTAbodies, lack of methodological agility and unnecessarybureaucracy resulting in belated identification of geneticvariants, inadequate description of clinical trial designs,incomplete representation of patient experiences, etc.[52]. Additionally, the review of scientific articles con-taining evidence to help guide decision-making about in-surance coverage for Alzheimer’s disease diagnostictests, conducted by the Institute for Clinical and Eco-nomic Review (ICER) and Policy Development Group(PDG), showed that all studies without exclusion failedto provide convincing evidence that payers could use toshowcase improved outcomes [53]. None of these stud-ies established analytic validity by capturing action basedupon diagnosis, health outcomes (e.g., cognitive/functiondecline), societal outcomes (e.g., cost-effectiveness), ortechnical efficacy [53].As a result, a great number of companion MDx devel-

oped separately from therapeutics did not receive widepayer acceptance due to poorly established links betweentesting, therapeutic interventions, and health outcomes(e.g., tests to estimate warfarin dosage or CYP2C19 as-says to stratify clopidogrel-eligible subpopulation) [54].Additionally, according to the literature review by Paciand Ibaretta [55], 27 % of the assessed Dx tests failed todemonstrate favorable and univocal cost-effectivenessevidence compared to the standard of care [55].Of the evaluation frameworks commonly cited by the

third-party payers, each has a different approach to tech-nology assessment, therefore creating inconsistencies inreimbursement decisions (Table 2).The table illustrates how conflicting the coverage deci-

sions of leading US insurers may be with respect to thesame molecular diagnostics. While some molecular as-says have been included in the coverage list by all major

Table 1 Examples of costs included in a typical cost-effectiveness analysis based on the perspective of the analysis (adapted fromTask Force on Community Preventive Services [50])

Cost Perspective

Societal Insurer/payer Employer Patient

Direct medical Yes Yesa Yesa Yesb

Direct non-medical (e.g., transportation, day care) Yes No No Yes

Indirect (e.g., time lost from work) Yesc No Yesc Yesc

Intangible (e.g., pain and suffering) Yesc No No Yesc

aCovered paymentsbOut-of-pocket paymentscIf not incorporated in the effect measure

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 6 of 12

payers as of 2010 (e.g., Oncotype Dx™, BRACAnalysis™,and the cobas® KRAS mutation test), other tests haveonly got positive coverage decisions by one or twopayers (e.g., like AlloMap™, MammaPrint™, and OVA1™).This could be partially explained by the fact that in theUSA, some payers use up to seven assessment frame-works to reason their reimbursement decisions (e.g.,BCBS TEC, Hayes, ECRI, EGAPP, ICER, USPSTF, andUp-To-Date), while others give preferences to only oneor two. Such discrepancies in assessment approachesand coverage policies create unnecessary confusion formanufacturers misguiding physicians and patients onthe clinical value of molecular diagnostic tests.The lack of consistency in value demarcation has

pushed the ICER to release a “value framework” draft,which is set to streamline the assessment processes frompayer perspective and to facilitate more structured dis-cussions between various stakeholders around value ofnew health technologies.The framework suggests that apart from comparative

clinical effectiveness measuring the degree of the com-parative net health benefit versus costs mentioned above,the assessment process must also incorporate contextualconsiderations enhancing the value of diagnostics andfavoring more informed reimbursement decisions. Forexample, it makes a huge difference for public and pri-vate insurers, whether they are dealing with an innova-tive technology or there is already an alternative existingon the market (e.g., new biomarkers and different mech-anisms of action) [57]. Amidst other contextual consid-erations, which increase the value of a MDx in the eyesof payers, are the following:

▪ High severity of the condition (e.g.,neurodegeneration and oncology);▪ High vulnerability of the population (e.g., pregnantwomen and children);▪ Ability to expand the eligible population for treatment(e.g., higher sensitivity/specificity);▪ Ability to improve the delivery system (e.g.,preparation, storage, or delivery of the therapy);

▪ Ability to decrease the misuse or overuse of therapies(e.g., precise drug dosage, over/under diagnosis, over-/undertreatment);▪ Low risk of pushback from precedents (e.g., otherpayer coverage decisions);▪ Low risk of pushback from clinicians and providergroups if not preferred (e.g., high complexity of tests),etc. [57, 58]

The ICER framework is one of the first attempts toconvince payers to integrate cost-effectiveness thinkinginto the way that they actually conduct their assessmentsin a more coherent way, and to encourage developmentof new policy tools to attain high health system value inconjunction with high clinical care value [58]. However,without clear guidelines on quantification of value ofdiagnostic tests, it may hardly occur that reimbursementauthorities arrive at a consistency in the coverage deci-sions on their own.

State perspectiveIn light of the nonhomogeneous global economic andpolitical environment, significant differences in HTAand reimbursement are observed at a state level, prede-fining the diversity of reimbursement and budgetaryallocation practices. Some countries, like the UK, rely oncost-utility analysis in their reimbursement decisions byevaluating the cost of interventions versus the obtainedhealth benefit (e.g., QALY). Other countries, like Franceand Germany, consider clinical added value with subse-quent “value for money” pricing debates.High variability of approaches to assess cost-effectiveness

of molecular diagnostics can be exemplified by EGFR test-ing before Iressa® (gefitinib) trial. While the MDx pro-ducer’s submission estimated cost-effectiveness of the testat GBP 23,615 per QALY, the assessment conducted byNICE demonstrated GBP 35,700 per QALY and the studyby SMC measured cost-effectiveness as GBP 154,022 perQALY [54]. This signifies the need for better modeling toolsand coherent methodological frameworks to assess eco-nomic outcomes of MDx strategies.

Table 2 Coverage inconsistencies in molecular diagnostics (adapted from Gustavsen et al. [56]

Innovative test example Positive coverage policies

Aetna Regional CMS Cigna Regional BCBS

AlloMap™ X

OncoType Dx™ (breast cancer) X X X X

MammaPrint™ X

BRACAnalysis™ X X X X

OVA1™ X X

KRAS (ovarian cancer) X X X X

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 7 of 12

The diversity in HTA systems is further complicated bycentralized versus decentralized reimbursement decision-making. In most of countries, coverage for prescriptionmedicines is done at the national level, while diagnosticmanufacturers have to deal with local payers in order toget reimbursement for their clinical diagnostic (CDx)tests. For example, despite of the fact that trastuzumab isreimbursed in most of European member states, itsHER2/neu companion diagnostic test is publicly funded inthe UK, Germany, and Italy, while in Spain, it is coveredby a therapeutic partner [59].Countries employ miscellaneous analytical frameworks

to guide their evaluations assessing multiple criteria likeclinical effectiveness, safety, costs of a technology versusits benefits, etc. (Table 3).While it is a policy in question that predetermines

the use of a particular analytical framework, the vastmajority of public assessment models account for clin-ical and patient benefit in the first place. Moreover, theevolving role of societal perspective in the evaluationprocess is shaping the assessment mechanisms resultingin the fact that today, some countries start consideringcosts and benefits outside the health sector as add-itional criteria for assessing value. For example, theWorking Group 4 report introduced by Busse et al. [61]supported the idea that assessment should go beyondsafety, efficacy/effectiveness, and economic aspects ac-counting for other meaningful outcomes, like thefollowing:

� Psychological/social/ethical outcomes (e.g., compliance,acceptance, satisfaction, demand, preferences, andinformation/patient advice requirements)

� Organizational/professional outcomes (e.g.,utilization of service, change in the treatmentlocation, change in length of hospital stay, change in

required personnel, material inputs, and trainingrequirements) [61]

To sum up, the heterogeneity of approaches to assessvalue of an innovative health technology, includingMDx, create significant market access hurdles forinnovative Dx tests. Therefore, joined efforts amongvarious stakeholders are required to shape the way as-sessment is carried out in different healthcare systems.This could be achieved by means of a common languageand terminology, uniform HTA assessment frameworks,more transparent and comprehensive reporting of thefindings, etc.

Case studiesIn the history of personalized medicine, there have beennumerous examples of molecular tests receiving positiveand negative reimbursement decisions by private andpublic payers in the USA and Europe. Below are severalcase studies from miscellaneous therapeutic areas whichbriefly describe coverage decisions for the selected mo-lecular diagnostic tests.

Oncology—breast cancerDako’s immunohistochemistry assay HercepTest™ is oneof the brightest examples of the successful application ofmolecular diagnostics in oncology by targeting HER2/neu overexpression in metastatic breast cancer. The testenabled stratification of breast cancer patients into eli-gible and non-eligible for Roche/Genentech’s monoclo-nal antibody therapeutic Herceptin® (trastuzumab) basedon the prognostic biomarker HER2/neu presented in20–30 % of breast cancer patients [62, 63]. The reim-bursement decisions for the HER2/neu companiondiagnostic were largely based on the results of RCTs,cost-effectiveness, and the ability to link the test toQALYs gained from treating the stratified subpopulationreceiving trastuzumab.While Herceptin® is commonly reimbursed in most of

the EU member states, no single test has been acceptedglobally as the gold standard for measuring HER2/neuoverexpression since 1998. Thus, in the UK, Germany,and Italy, HER2 test is covered by the state, while inSpain, it is an Rx partner who pays for testing. In France,HER2 test was authorized in 2000 but received positivereimbursement decision only in 2007 [64].

Oncology—colorectal cancerApproximately 30–50 % of colorectal tumors are associ-ated with an abnormal KRAS gene, signifying that nearlyhalf of patients with colorectal cancer (CRC) mightrespond to anti-epidermal growth factor receptor (EGFR)treatment, and the other half might not [65]. Two mono-clonal antibodies, Erbitux® (cetuximab) and Vectibix®

Table 3 Criteria for health technology assessment inEurope (adapted from Sorenson et al. [60])

Criteria AT BE CH DE FI FR NL NO SE UK

Clinical benefit X X X X X X X X X X

Patient benefit X X X X X X X X X X

Cost-effectiveness X X X X X X X

Budget impact X X X X X X

Innovative characteristics X X X X X

Availability of alternatives X X X X

Equity considerations X X X

Public health impact X

R&D X

Reproduced with permission from the European Observatory on HealthSystems and PoliciesAT Austria, BE Belgium, CH Switzerland, DE Denmark, FI Finland, FR France, NLNetherlands, NO Norway, SE Sweden, UK United Kingdom

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 8 of 12

(panitumumab), have demonstrated favorable survival im-pact in population with KRAS wild-type CRC [66, 67].Hence, KRAS mutation testing is currently used in clinicalpractice to assist in identifying eligible patients for anti-body treatment.The positive reimbursement decisions of the majority

of American and European payers were primarily basedon the strong linkage between the clinical utility of thetest and patient outcomes derived from the post factumanalysis of cetuximab and panitumumab clinical trialdata [68-70], cost-effectiveness [71], and inclusion of thetest in clinical practice guidelines [72]. This is one of thefirst examples of diagnostic companies receiving widepayer acceptance based solely on retrospective archivedsamples data without conducting heavy and costly RCTs.Other examples of successful reimbursement of mo-

lecular diagnostics in oncology include Oncotype DX®,MammaPrint®, and OncoVue® (breast cancer); EGFRmutation assays (non-small cell lung cancer); BCR-ABLtesting (chronic myelogenous leukemia); and others.

HIVThe HLA-B*5701 molecule is associated with hypersensi-tivity to the antiretroviral drug abacavir; hence, a positivetest allows physicians to use the drug in a much largerpopulation due to the assay’s ability to identify patientsthat are at risk of developing severe adverse events [73].The reimbursement decisions for the HLA-B*5701

testing were primarily based on data from RCTs andretrospective studies, cost-effectiveness, and tests’ abilityto identify broader population of patients with highdegree of sensitivity and specificity. The cost of the testwas justified by clear clinical utility (prospective andretrospective studies showing improved health outcomes[74]), cost benefits (including management of theadverse event [73]), and cost-effectiveness (e.g., incre-mental cost-effectiveness ratio per hypersensitivity reac-tion avoided [75]). Additionally, inclusion of abacavirinto the treatment guidelines along with the companiondiagnostic has reinforced reimbursement of HLA-B*5701 test by private and public payers [76].

Hepatitis CViral load monitoring (VLM) is a quantitative PCR orbDNA test used by physicians to confirm the presenceof HCV and to measure the patient’s viral load withhepatitis C. Despite the fact that VLM does not predictthe development of cirrhosis or liver failure, like theaforementioned HIV testing tools, it shows the likeli-hood of patient’s response to treatment.The ability to predict therapeutic outcomes, avoiding

costs associated with treatment of non-respondents andminimizing adverse effects associated with the intervention,

laid the basis for the test to be include in the reimburse-ment lists by payers in the USA and Europe [77].

CardiovascularCardioDX’ Corus® CAD is a multi-analyte gene expres-sion assay introduced as a less invasive way to identifyobstructive coronary artery disease (CAD) and assist pri-mary care physicians and cardiologists to understandwhether the symptoms of cardiovascular disease in non-diabetic patients are caused by CAD.Seeking for wide payer acceptance, CardioDX has

invested heavily into the three costly clinical utility studies(PREDICT, COMPASS [78], and IMPACT-PCP [79]) andreceived Medicare reimbursement via the PalmettoMolDx program based on reduced costs by avoiding cor-onary angiography and other invasive assays [80]. How-ever, of the 10 largest private payers in the USA, AetnaHealth was the only organization that included the testinto its coverage list, while the rest of payers rejectedCorus® CAD because it had a poorly established link be-tween clinical decision-making and patient outcomes [81].This case study signifies that massive prospective stud-

ies, which lay the ground for clinical utility, are import-ant, but not the only critical point in a reimbursementdecision; therefore, companies have to carefully accountfor other payer requirements with regard to economic,clinical, and social value of molecular diagnostics, inorder to get reimbursement and wider test acceptance.

ConclusionsInnovative molecular diagnostics are becoming an essen-tial part of disease management and therapy, helpingphysicians to stratify patient cohorts, choose more ap-propriate drug regimen, avoid adverse events, facilitatetherapeutic monitoring, and define the predisposition toa disease. Notwithstanding their importance, MDx arestill frequently perceived as “additional costs” and getmuch lower reimbursement rates and profit marginsthan therapeutics. However, in the nearest term, the newgeneration molecular assays are expected to alter thisstatus quo, either by becoming “companion diagnostics”or by demonstrating substantial value, like cost reduc-tion through targeted treatment, reduction in the num-ber of side effects, etc.Due to the polysemic nature of “value,” there are

many ways to assess MDx depending on which frame-work is used and from whose perspective the technol-ogy is being evaluated. Thereby, it is recommended forindustry to facilitate discussions with various stake-holders throughout the assessment process in order toarrive at a consensus about the depth of evidence re-quired for positive regulatory/reimbursement decisionsand clinical adoption, gazing beyond accuracy andfeasibility data. It is also recommended to develop clear

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 9 of 12

guidelines on quantification of value of diagnostic teststo avoid inconsistency in the coverage decisions bypublic and private payers.The further lines of our research will address the follow-

ing topics: (1) the impact of social and ethical parametersof value on reimbursement and test acceptance in light ofthe evolving “value-based healthcare” delivery practices;(2) the evolving innovative trial designs driven by smallerMDx-based studies with stratified patient cohorts and ar-chived specimen, contributing to level I evidence and re-placing massive RCTs; and (3) investigation into theheterogeneity of the state-level HTA frameworks.

AbbreviationsCDx: clinical diagnostic; Dx: diagnostic; MDx: molecular diagnostic tests.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsIA suggested the idea and wrote the article. RVB did the analysis of the data,editing of the manuscript, and medical advising, and performed the finalarticle drafting. Both authors read and approved the final manuscript.

Authors’ informationIA is a Market Access Consultant at Unicorn in Zhytomyr, Ukraine.RVB, M.D., Ph.D. is a medical doctor in the Clinical Hospital “Pheophania” ofthe State Affairs Department, researcher of the Interferon Department ofZabolotny Institute of Microbiology and Virology, National Academy ofSciences of Ukraine, and National Representative of the EuropeanAssociation for Predictive, Preventive and Personalized Medicine (EPMA) inUkraine.

AcknowledgementsWe acknowledge the EPMA Journal editorial team and BioMed Central teamfor the opportunity to publish this work.

Author details1Strategic Market Intelligence Dep., Unicorn, P.O.B. 91, Zhytomyr 10020,Ukraine. 2Clinical Hospital “Pheophania” of State Affairs Department,Zabolotny Str., 21, Kyiv 03680, Ukraine. 3Zabolotny Institute of Microbiologyand Virology, National Academy of Sciences of Ukraine, Zabolotny Str., 154,Kyiv 03680, Ukraine.

Received: 22 July 2015 Accepted: 8 September 2015

References1. Golubnitschaja O, Costigliola V, EPMA. General report & recommendations

in predictive, preventive and personalised medicine 2012: white paper ofthe European Association for Predictive, Preventive and PersonalisedMedicine. EPMA J. 2012;3(1):14. doi:10.1186/1878-5085-3-14.

2. Grech G, Zhan X, Yoo BC, Bubnov R, Hagan S, Danesi R, et al. EPMA positionpaper in cancer: current overview and future perspectives. EPMA J. 2015;6:9.doi:10.1186/s13167-015-0030-6.

3. Chow IP, Bryant T, Swanson S, Schoeninger B. Dissecting the value ofcompanion diagnostics. IMS Consulting Group. 2013. Accessed from http://www.imshealth.com/deployedfiles/imshealth/Global/Content/Healthcare/IMS%20Consulting/Our%20Latest%20Thinking/Q1_2013_IMSCG_Market_Access_Conference_Dissecting_The_Value_of_CDx_07_Mar2013.pdf on May18, 2015.

4. Carson J. Molecular methods rapidly gain ground in infectious diseasediagnostics. Frost & Sullivan. 2014. Accessed from http://ww2.frost.com/news/press-releases/molecular-methodsrapidly-gain-ground-infectious-disease-diagnostics-says-frost-sullivan/ on May 17, 2015.

5. NCBI. GTR: Genetic Testing Registry. NCBI. 2015. Accessed from http://www.ncbi.nlm.nih.gov/gtr/ on May 29, 2015.

6. Di Ruffano F, Hyde L, McCaffery CJ, Bossuyt KJ, Patrick MM, Deeks JJ.Assessing the value of diagnostic tests: a framework for designing andevaluating trials. BMJ. 2012;344, e686.

7. Lord SJ, Irwig L, Simes RJ. When is measuring sensitivity and specificitysufficient to evaluate a diagnostic test, and when do we need randomizedtrials? Ann Intern Med. 2006;144(11):850–5. doi:10.7326/0003-4819-144-11-200606060-00011.

8. Forsman RW. Why is the laboratory an afterthought for managed careorganizations? Clin Chem. 1996;42(5):813–6. Review.

9. Hanson C, Plumhoff E. Test utilization and the clinical laboratory. MayoClinic. 2012. Accessed from http://www.mayomedicallaboratories.com/articles/communique/2012/05.html on May 23, 2015.

10. Coamey J. MDx and MDs: is a dose of knowledge the prescription foradoption? CAHG. 2010. Accessed from http://www.healthtech.com/uploadedFiles/Conferences/White_Papers/MMTC/CAHG_LandmarkPhysicianStudy.pdf on May 18, 2015.

11. Drucker E, Krapfenbauer K. Pitfalls and limitations in translation frombiomarker discovery to clinical utility in predictive and personalisedmedicine. EPMA J. 2013;4(1):7. doi:10.1186/1878-5085-4-7.

12. Zelin J, Garrett N, Saunders J, Warburton F, Anderson J, Moir K, et al. NorthEast London Sexual Health Network Research Consortium. An evaluation ofthe performance of OraQuick® ADVANCE Rapid HIV-1/2 Test in a high-riskpopulation attending genitourinary medicine clinics in East London, UK. IntJ STD AIDS. 2008;19(10):665–7. doi:10.1258/ijsa.2008.008132.

13. Caragher TE, Fernandez BB, Jacobs FL, Barr LA. Evaluation of quantitativecardiac biomarker point-of-care testing in the emergency department.J Emerg Med. 2002;22(1):1–7.

14. McCord J, Nowak RM, McCullough PA, Foreback C, Borzak S, Tokarski G, etal. Ninety-minute exclusion of acute myocardial infarction by use ofquantitative point-of-care testing of myoglobin and troponin I. Circulation.2001;104(13):1483–8.

15. Jørgensen OD, Kronborg O, Fenge C. A randomised study of screening forcolorectal cancer using faecal occult blood testing: results after 13 yearsand seven biennial screening rounds. Gut. 2002;50(1):29–32.

16. Forrest GN, Roghmann MC, Toombs LS, Johnson JK, Weekes E, Lincalis DP, et al.Peptide nucleic acid fluorescent in situ hybridization for hospital-acquiredenterococcal bacteremia: delivering earlier effective antimicrobial therapy.Antimicrob Agents Chemother. 2008;52(10):3558–63. doi:10.1128/AAC.00283-08.

17. Veenstra DL. The cost-effectiveness of warfarin pharmacogenomics. JThromb Haemost. 2007;5(9):1974–5. doi:10.1111/j.1538-7836.2007.02699.x.

18. Carlson B. KRAS testing: optimizing cancer therapy. Biotechnol Healthc.2009;6(5):7–9.

19. Asad J, Jacobson AF, Estabrook A, Smith SR, Boolbol SK, Feldman SM, et al.Does oncotype DX recurrence score affect the management of patientswith early-stage breast cancer? Am J Surg. 2008;196(4):527–9. doi:10.1016/j.amjsurg.2008.06.021.

20. Rich PA, Shaefer CF, Parkin CG, Edelman S. Using a quantitative measure ofdiabetes risk in clinical practice to target and maximize diabetes preventioninterventions. Clin Diab. 2013;31(2):82–9. doi:10.2337/diaclin.31.2.82.

21. Schneider KI, Schmidtke J. Patient compliance based on genetic medicine: aliterature review. J Comm Genet. 2014;5(1):31–48. doi:10.1007/s12687-013-0160-2.

22. PMC. Personalized medicine by the numbers. [Infographic]. PMC. 2014. Accessedfrom http://www.personalizedmedicinecoalition.org/Userfiles/PMCCorporate/file/pmc_personalized_medicine_by_the_numbers.pdf on June 2, 2015.

23. Falconi A, Lopes G, Parker JL. Biomarkers and receptor targeted therapiesreduce clinical trial risk in non-small-cell lung cancer. J Thorac Oncol.2014;9(2):163–9. doi:10.1097/JTO.0000000000000075.

24. Fitzal F, Filipits M, Rudas M, Greil R, Dietze O, Samonigg H, et al. Thegenomic expression test EndoPredict is a prognostic tool for identifying riskof local recurrence in postmenopausal endocrine receptor-positive,her2neu-negative breast cancer patients randomised within the prospectiveABCSG 8 trial. Br J Cancer. 2015;112(8):1405–10. doi:10.1038/bjc.2015.98.

25. Garau M, Towse A, Garrison L, Housman L, Ossa D. Can and should valuebased pricing be applied to molecular diagnostics? 2012. Accessed fromhttps://www.ohe.org/publications/canand-should-value-based-pricing-be-applied-molecular-diagnostics on May 21, 2015.

26. Lepik KJ, Harrigan PR, Barrios R, Milan D, Yip B, Chan K, et al. The impact ofinterventions to improve HLA-B*5701 allele screening to reduce risk ofabacavir hypersensitivity reaction. AIDS 2010 - XVIII International AIDSConference: Abstract no.WEPE0144. Accessed from https://www.iasociety.org/Abstracts/A200735296.aspx on May 25, 2015.

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 10 of 12

27. Armour AA, Watkins CL. The challenge of targeting EGFR: experience withgefitinib in nonsmall cell lung cancer. Eur Respir Rev. 2010;19(117):186–96.doi:10.1183/09059180.00005110.

28. Spackman E, Rice S, Norman G, Suh DC, Eastwood A, Palmer S. Trastuzumabfor the treatment of HER2-positive metastatic gastric cancer.Pharmacoeconomics. 2013;31(3):185–94. doi:10.1007/s40273-013-0023-z.

29. Porter ME. Value-based health care delivery. [Video] HBS. 2012. Accessedfrom https://www.youtube.com/watch?v=DRkhppxZzL0 on May 29, 2015.

30. Porter ME. Value-based health care delivery. [Slides] HBS. 2012. Accessedfrom http://www.hbs.edu/healthcare/Documents/2012%2003%2007%20SUT%20HCI%20presentation.pdf on May 29, 2015.

31. Cosgrove T. Value-based health care is inevitable and that’s good. HarvardBusiness Review. 2013. Accessed from https://hbr.org/2013/09/value-based-health-care-is-inevitable-and-thatsgood/on May 19, 2015.

32. Catawba Valley Medical Center. CHESS And Catawba Valley Medical Center(CVMC) partner to adapt value-based healthcare delivery model. CatawbaValley Medical Center. 2015. Accessed from https://catawbavalleymedical.org/chess-catawba-valley-medical-center-cvmc-partner-adapt-value-based-healthcare-delivery-model/ on May 17, 2015.

33. MHA Task Force on Future Health Insurance Markets. Health insuranceexchanges and the road to value-based payment in Michigan. MHHA. 2014.Accessed from http://www.mha.org/documents/tf_fhim_final_report_061814.pdf on May 27, 2015.

34. Carlson B. A truer way to measure value of diagnostic tests. BiotechnolHealthc. 2012;9(4):34–5.

35. Fryback DG, Thornbury JR. The efficacy of diagnostic imaging. Med DecisMak. 1991;11(2):88–94. doi:10.1177/0272989x9101100203.

36. Silverstein MD, Boland BJ. Conceptual framework for evaluating laboratorytests: case-finding in ambulatory patients. Clin Chem. 1994;40(8):1621–7.

37. Mackenzie R, Dixon AK. Measuring the effects of imaging: an evaluativeframework. Clin Radiol. 1995;50(8):513–8. doi:10.1016/S0009-9260(05)83184-8.

38. Pearson SD, Knudsen AB, Scherer RW, Weissberg J, Gazelle GS.Assessing the comparative effectiveness of a diagnostic technology:CT colonography. Health Aff (Millwood). 2008;27(6):1503–14. doi:10.1377/hlthaff.27.6.1503.

39. Harris RP, Helfand M, Woolf SH, Lohr KN, Mulrow CD, Teutsch SM, et al.Third US Preventive Services Task Force. Current methods of the U.S.Preventive Services Task Force. Am J Prev Med. 2001;20(3 Suppl):21–35.

40. Trikalinos TA, Siebert U, Lau J. Decision-analytic modeling to evaluatebenefits and harms of medical tests: uses and limitations. Med Decis Mak.2009;29(5):E22–9. doi:10.1177/0272989x09345022.

41. Schünemann HJ, Oxman AD, Brozek J, Glasziou P, Jaeschke R, Vist GE, et al.Grading quality of evidence and strength of recommendations fordiagnostic tests and strategies. BMJ. 2008;336(7653):1106–10. doi:10.1136/bmj.39500.677199.AE.

42. Hayes DF, Bast RC, Desch CE, Fritsche Jr H, Kemeny NE, Jessup JM, et al.Tumor marker utility grading system: a framework to evaluate clinical utilityof tumor markers. J Natl Cancer Inst. 1996;88(20):1456–66. doi:10.1093/jnci/88.20.1456. Review.

43. Simon RM, Paik S, Hayes DF. Use of archived specimens in evaluation ofprognostic and predictive biomarkers. JNCI J National Cancer Institute.2009;101(21):1446–52. doi:10.1093/jnci/djp335.

44. National Academy of Sciences. Generating evidence for genomic diagnostictest development: workshop summary. Institute of Medicine (US) Roundtableon Translating Genomic-Based Research for Health. 2011. Accessed fromhttp://www.ncbi.nlm.nih.gov/books/NBK4119/ on July 2, 2015.

45. Wald N, Cuckle H. Reporting the assessment of screening and diagnostictests. Br J Obstet Gynaecol. 1989;96(4):389–96. doi:10.1111/j.1471-0528.1989.tb02411.x.

46. Haddow JE, Palomaki GE. An introduction to assessing genomic screeningand diagnostic tests. Nutr Today. 2011;46(4):162–8.

47. Bossuyt PM, McCaffery K. Additional patient outcomes and pathways inevaluations of testing. Med Decis Making. 2009;29(5):E30–8. doi:10.1177/0272989X09347013.

48. Lee DW, Neumann PJ, Rizzo JA. Understanding the medical andnonmedical value of diagnostic testing. Value Health. 2010;13(2):310–4.doi:10.1111/j.1524-4733.2009.00597.x.

49. Kopits IM, Chen C, Roberts JS, Uhlmann W, Green RC. Willingness to pay forgenetic testing for Alzheimer’s disease: a measure of personal utility. GenetTest Mol Biomarkers. 2011;15(12):871–5. doi:10.1089/gtmb.2011.0028.

50. Task Force on Community Preventive Services. The guide to communitypreventive services—what works to promote health? Oxford UniversityPress. 2005. Accessed from http://www.thecommunityguide.org/library/book/Front-Matter.pdf on June 5, 2015.

51. Grosse SD, Wordsworth S, Payne K. Economic methods for valuing theoutcomes of genetic testing: beyond cost-effectiveness analysis. Genet Med.2008;10(9):648–54.

52. The Deerfield Institute. The Deerfield Institute—EuropaBio report onregulatory and scientific advice for small and medium enterprises. 2015.Accessed from http://www.europabio.org/sites/default/files/report/deerfield_europabio_survey_regulatory_and_hta_advice_for_smes_-_march_2015.pdf on June 7, 2015.

53. Foster NL, Hackett JSM, White G, Chenevert S, Svarvar P, Bain L, et al.Justifying reimbursement for Alzheimer’s diagnostics and treatments:seeking alignment on evidence. Alzheimers Dement. 2014;10(4):503–8.doi:10.1016/j.jalz.2014.05.003.

54. Faulkner E, Annemans L, Garrison L, Helfand M, Holtorf AP, Hornberger J, etal. Challenges in the development and reimbursement of personalizedmedicine—payer and manufacturer perspectives and implications for healtheconomics and outcomes research: a report of the ISPOR PersonalizedMedicine Special Interest Group. Value Health. 2012;15(8):1162–71.doi:10.1016/j.jval.2012.05.006.

55. Paci D, Ibarreta D. Economic and cost-effectiveness considerations forpharmacogenetics tests: an integral part of translational research andinnovation uptake in personalized medicine. Current Pharmacogenomics,Personalized Med. 2009;7(4):284–96. doi:10.2174/187569209790112355.

56. Gustavsen G, Phillips K, Pothier K. The reimbursement landscape for noveldiagnostics: current limitations, real-world impact, and proposed solutions:health advances. 2010. Accessed from https://www.bio.org/sites/default/files/Health_Advances%26BIO_Novel_Diagnostics_Reimburs_20110103.pdfon June 9, 2015.

57. Pearson SD. Assessing the “value” of tests and treatments in the US healthcare system. [Video] Philip R. Lee Institute for Health Policy Studies. 2014.Accessed from https://www.youtube.com/watch?v=yFZ2oUUrovE on May29, 2015.

58. Pearson SD. A framework for payer assessment of the value of newtechnologies: a US approach (seminar briefing). Office of Health Economics.2015. Accessed from https://www.ohe.org/publications/framework-payer-assessment-value-new-technologies-us-approach on June 1, 2015.

59. Akhmetov IR, Ramaswamy R, Akhmetov IR, Thimmaraju P. Market accessadvancements and challenges in “drug-companion diagnostic test”co-development in Europe. J Pers Med. 2015;5(2):213–28. doi:10.3390/jpm5020213.

60. Sorenson C, Drummond M, Kanavos P. Ensuring value for money in healthcare. The role of health technology assessment in the European Union.European Observatory on Health Systems and Policies. 2008. Accessed fromhttp://www.euro.who.int/__data/assets/pdf_file/0011/98291/E91271.pdf onJune 7, 2015.

61. Busse R, Orvain J, Velasco M, Perleth M, Drummond M, Gürtner F, et al. Bestpractice in undertaking and reporting health technology assessments. Int JTechnol Assess Health Care. 2002;18(2):361–422.

62. Barrett C, Magee H, O’Toole D, Daly S, Jeffers M. Amplification of the HER2gene in breast cancers testing 2+ weak positive by HercepTestimmunohistochemistry: false‐positive or falsenegative immunohistochemistry?J Clin Pathol. 2007;60(6):690–3. doi:10.1136/jcp.2006.039602.

63. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, Bajamonde A, et al.Use of chemotherapy plus a monoclonal antibody against HER2 formetastatic breast cancer that overexpresses HER2. N Engl J Med.2001;344(11):783–92.

64. Garfield S. Advancing access to personalized medicine: a comparativeassessment of European reimbursement systems. PMC. 2011. Accessed fromhttp://www.personalizedmedicinecoalition.org/Userfiles/PMC-Corporate/file/pmc_bridgehead_issue_brief_european_reimbursement.pdf on May 23,2015.

65. Wilson PM, LaBonte MJ, Lenz HJ. Molecular markers in the treatment ofmetastatic colorectal cancer. Cancer J. 2010;16(3):262–72. doi:10.1097/PPO.0b013e3181e07738.

66. Amado RG, Wolf M, Peeters M, Van Cutsem E, Siena S, Freeman DJ, et al.Wild-type KRAS is required for panitumumab efficacy in patients withmetastatic colorectal cancer. J Clin Oncol. 2008;26(10):1626–34. doi:10.1200/jco.2007.14.7116.

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 11 of 12

67. Karapetis CS, Khambata-Ford S, Jonker DJ, O’Callaghan CJ, Tu D, Tebbutt NC,et al. K-ras mutations and benefit from cetuximab in advanced colorectalcancer. N Engl J Med. 2008;359(17):1757–65. doi:10.1056/NEJMoa0804385.

68. Rojas M, Yao SY, Lin YZ. Controlling epidermal growth factor (EGF)-stimulated Ras activation in intact cells by a cell-permeable peptidemimicking phosphorylated EGF receptor. J Biol Chem. 1996;271(44):27456–61.doi:10.1074/jbc.271.44.27456.

69. Van Cutsem E, Peeters M, Siena S, Humblet Y, Hendlisz A, Neyns B, et al.Open-label phase III trial of panitumumab plus best supportive carecompared with best supportive care alone in patients with chemotherapy-refractory metastatic colorectal cancer. Clin Oncol. 2007;25(13):1658–64.

70. Van Cutsem E, Nowacki M, Lang I, et al. Randomized phase III study ofirinotecan and 5-FU/FA with or without cetuximab in the first-linetreatment of patients with metastatic colorectal cancer (mCRC): theCRYSTAL trial. J Clin Oncol. 2007;25(Suppl):18s–s.

71. Westwood M, van Asselt T, Ramaekers B, Whiting P, Joore M, Armstrong N,et al. KRAS mutation testing of tumours in adults with metastatic colorectalcancer: a systematic review and cost-effectiveness analysis. Health TechnolAssess. 2014;18:62. doi:10.3310/hta18620.

72. Allegra CJ, Jessup JM, Somerfield MR, Hamilton SR, Hammond EH, Hayes DF,et al. American society of clinical oncology provisional clinical opinion:testing for KRAS gene mutations in patients with metastatic colorectalcarcinoma to predict response to anti-epidermal growth factor receptormonoclonal antibody therapy. J Clin Oncol. 2009;27(12):2091–6. doi:10.1200/JCO.2009.21.9170.

73. Mallal S, Phillips E, Carosi G, Molina JM, Workman C, Tomazic J, et al.PREDICT-1 study team. HLA-B*5701 screening for hypersensitivity toabacavir. N Engl J Med. 2008;358(6):568–79. doi:10.1056/NEJMoa0706135.

74. Ma JD, Lee KC, Kuo GM. HLA-B*5701 testing to predict abacavirhypersensitivity. PLoS Currents. 2010;2, RRN1203. doi:10.1371/currents.RRN1203.

75. Hughes DA, Vilar FJ, Ward CC, Alfirevic A, Park BK, Pirmohamed M. Cost-effectiveness analysis of HLA B*5701 genotyping in preventing abacavirhypersensitivity. Pharmacogenetics. 2004;14(6):335–42.

76. Aberg JA, Kaplan JELH, Emmanual P, Anderson JR, Stone VE, Oleske JM.Primary care guidelines for the management of persons infected withhuman immunodeficiency virus: 2009 update by the HIV MedicineAssociation of the Infectious Diseases Society of America. Clin Infect Dis.2009;49:651–81.

77. Towse A, Ossa D, Veenstra D, Carlson J, Garrison L. Understanding theeconomic value of molecular diagnostic tests: case studies and lessonslearned. JPM. 2013;3:288–305. doi:10.3390/jpm3040288.

78. CardioDx. Corus CAD clinical validity. CardioDx. 2015. Accessed from http://www.cardiodx.com/research-and-clinical-data/corus-cad-development-and-data/clinical-validity/ on June 7, 2015.

79. CardioDx. Study finds the Corus® CAD gene expression blood testinfluenced primary care decision-making in the assessment of patients withsuspected obstructive coronary artery disease. 2015. Accessed from http://www.cardiodx.com/media-and-events/press-releases/study-finds-the-corus-cad-gene-expression-blood-test-influenced-primary-care-decision-making-inthe-assessment-of-patients-with-suspected-obstructive-coronary-artery-disease/ on May 17, 2015.

80. Palmetto GBA. Approved gene testing (M00041, V8). Palmetto GBA. 2015.Accessed from http://www.palmettogba.com/palmetto/MolDX.nsf/DocsCat/MolDx%20Website~MolDx~Browse%20By%20Topic~Covered%20Tests~9BMLRK6738?open&navmenu=%7C%7C on May 28, 2015.

81. Malone B. The reimbursement outlook for molecular diagnostics—shiftingpolicies, market demands trace uncertain path. American Association forClinical Chemistry. 2015. Accessed from https://www.aacc.org/publications/cln/articles/2015/april/the-reimbursement-outlook-for-molecular-diagnosticson May 26, 2015.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

Akhmetov and Bubnov The EPMA Journal (2015) 6:19 Page 12 of 12